The Maintenance of Plasmids in Pathogenic Organisms

细菌中低拷贝数质粒的研究有两个主要原因。首先，它们在许多方面就像细胞内的小的、可缺性的染色体一样，因此是研究染色体复制和分离的容易处理的模型。其次，它们具有相当大的医学重要性。它们是在致病菌中传播抗生素耐药性的可传播因素，在某些情况下，它们是人类传染病中细菌感染毒性的决定因素。抗生素耐药性的蔓延有可能使抗生素治疗在未来几十年几乎毫无用处。此外，含有毒力质粒的致病菌越来越多地成为癌症患者死亡的最终原因，这些患者的免疫系统往往因疾病进展或化疗而受损。因此，了解这些质粒是如何在细菌种群中稳定维持的，是开发感染性疾病治疗和质粒传播修复新策略的第一步，这一点非常重要。我们特别感兴趣的机制，质粒使用，以确保其适当分离到子细胞。我们研究了一个被称为分割基因的元件家族（P1par家族），它负责分离几种类型的质粒，包括导致肠道疾病的沙门氏菌和志贺氏菌以及鼠疫耶尔森菌的毒力质粒；黑死病腺鼠疫的致病生物在每种情况下，我们已经证明分离是通过识别顺式作用位点parS（类似于着丝粒）和两个质粒编码蛋白ParA和ParB来实现的。ParB与parS特异性结合，而parS是一种atp酶，在分离过程中可能是移动质粒的马达。P1par家族的成员表现出独特的物种特异性。这一点很重要，因为否则，不同类型的质粒就会相互竞争，从而限制它们在自然界的传播。我们发现这些物种特异性存在于ParB蛋白和parS位点之间的一种新的相互作用中。这不是为ParB结合到位点提供能量的相互作用。相反，它是ParB n端与parS中称为B盒的短基序之间的特殊接触。通过改变B盒序列的一个碱基，我们可以改变系统的特异性，从一个物种到另一个物种。这种机制似乎是一种新型的dna -蛋白质识别，可能对蛋白质在其他潜在结合位点存在时如何在特定位点起作用具有广泛的意义。我们进一步探讨了P1par分隔元件家族成员的物种特异性与顺式作用分隔位点BoxB序列与ParB蛋白的接触有关的假设。利用最近发表的P1 ParB蛋白的晶体结构作为指导，我们已经能够精确定位ParB蛋白表面的位置，该位置指定了单个质粒DNA parS位点上物种决定因子的识别。通过改变ParB蛋白中的单个氨基酸，我们已经能够将其识别特异性完全转换为不同物种的识别特异性。我们已经纯化了这些改变的蛋白质，并且已经能够证明它们改变的特异性并不存在于DNA结合能力的改变中。相反，BoxB DNA位点似乎是质粒分离分割复合体的特定激活剂。这对划分模型具有重要的意义。我们目前正通过荧光显微技术在活细胞中研究质粒DNA的分离来验证这些结论。此外，我们还与生物分子结构组（结晶学实验室，CCR）的纪新华博士合作，建立了改变相互作用的分子结构模型。结果与遗传观察结果吻合良好，并证实单个DNA碱基-氨基酸侧链相互作用是确定这些系统特异性的关键。我们开发的自动数据收集和分析活细菌细胞的荧光灶，使我们有机会重新访问P1质粒在大肠杆菌中的行为。研究结果令人惊讶，并表明目前关于这一主题的许多文献都是错误的。人们声称P1和相关的质粒类型在细胞中心复制并分离到细胞四分之一位置，准备在细胞分裂时处于子细胞的中心。因此，重点是寻找与隔离有关的宿主地点。现在，我们发现细胞内的固定位置没有明显的偏好。多个质粒病灶的平均位置彼此之间以及与细胞极点之间的间隔是均匀的。在时间推移和时间戳显微镜中分离过程的动力学揭示了质粒焦点相互远离和靠近的相当数量的运动。因此，种群和时间平均分布是均匀的，但单个细胞的模式变化很大。复制和分离几乎可以在细胞的任何位置发生。此外，在人群中观察到相似数量的焦点分裂和配对事件。这暗示了一种分裂模型，在这种模型中，质粒配对要么是由于复制的结果，要么是由于相当频繁的偶遇。然后它们沿着细胞轴迅速远离彼此，保持拷贝的动态分布，近似于长细胞轴上的均匀分布。这些观察结果对质粒物理分离的机制具有重要意义。今年，我们开发了一种质粒分离的数学模拟，类似于被称为蒙特卡洛模拟的模型类型。这个模拟是从我们提出的用来解释实验观察的分离规则推导出来的。模拟产生的质粒分离保真度和质粒拷贝在细胞内的空间分布在所有细节上都与实验数据非常相似。模拟大大增加了我们的信心，从我们的研究中出现的质粒分离的新画面是正确的。

Low copy number plasmids in bacteria are of interest for two principle reasons. First, they act in many ways like small, dispensable chromosomes within the cell, and are therefore tractable models for the study of chromosome replication and segregation. Second, they are of considerable medical importance. They are the transmissible elements that spread antibiotic resistance among pathogenic bacteria and in some cases, are the determinants of the virulence of bacterial infection in human infectious disease. The spread of antibiotic resistance threatens to make antibiotic therapy virtually useless in the next few decades. In addition, pathogenic bacteria containing virulence plasmids are increasingly the ultimate cause of death of cancer patients whose immune systems are often compromised by disease progression or chemotherapy. It is therefore of importance to try to understand how these plasmids are stably maintained in the bacterial population as a first step toward developing novel strategies for infectious disease therapy and remediation of plasmid spread. We are particularly interested in the mechanisms that plasmids use to ensure their proper segregation to daughter cells. We study a family of elements known as partition genes (the P1par family), that are responsible for the segregation of several types of plasmid including the virulence plasmids of Salmonella and Shigella species responsible for enteric disease, and of Yersinia pestis; the causative organism for bubonic plague. In each case, we have shown that segregation is achieved by recognition of a cis-acting site parS, analogous to a centromere, and two plasmid encoded proteins, ParA and ParB.. ParB binds specifically to parS and ParA is an ATPase that may be a motor for moving the plasmid during segregation. Members of the P1par family show unique species specificities. This is important, because, otherwise, plasmids of different types would compete with each other, limiting their spread in nature. We have discovered that these species specificities reside in a novel interaction between the ParB protein and the parS site. This is not the interaction that provides the energy for ParB binding to the site. Rather, it is a special contact between the ParB N-terminus and a short motif in parS termed the B box. By changing the B box sequence by as little as one base, we can change the specificity of the system from one species to another. This mechanism appears to be a novel type of DNA-protein recognition that may have broad implications for how proteins act at a specific site when other potential binding sites exist. We have further explored the hypothesis that the contact between the BoxB sequences in the cis-acting partition site and the ParB protein are responsible for the species specificity of members of the P1par family of partition elements. Using the recently published crystal structure of the P1 ParB protein as a guide, we have been able to pinpoint the position on the ParB protein surface that specifies recognition of species determinants on the individual plasmid DNA parS sites. By changing a single amino acid in the ParB protein we have been able to switch its recognition specificity completely to that of a different species. We have purified these altered proteins and have been able to show that thier altered specificity does not reside in a change in DNA binding ability. Rather, it appears that the BoxB DNA site acts as a specific activator of the partition complex for plasmid segregation. This has important implications for models of partition. We are currently testing these conclusions by studying the segregation of the plasmid DNA by fluorescence photomicroscopy in living cells. In addition, we have carried out a collaboration with Dr. Xinhua Ji of the Biomolecular Structure Section (Chystallography Laboratory, CCR), to model the molecular structures of the altered interactions. The results fit well with the genetic observations and confirm that a single DNA base-amino acid side chain interaction is key to determining the specificity of these systems. Our development of automated data collection and analysis of fluorescent foci in living bacterial cells have given us an opportunity to re-visit the behavior of the P1 plasmid in E. coli. The results have been surprising, and suggest that much of the current literature on this subject is wrong. It has been claimed that P1, and related plasmid types, replicate in the cell center and segregate to the cell quarter positions, ready to be at the center of the daughter cells at cell division. Thus, the emphasis has been on searching for host sites involved in segregation. Now, we find that there is no clear preference for fixed positions within the cell. The mean positions of multiple plasmid foci are evenly spaced from each other and from the cell poles. The dynamics of the segregation process in time-lapse and time-stamp microscopy reveals a considerable amount of movement of plasmid foci both away from and toward each other. Thus, the population and time-averaged distributions are even, but individual cell patterns vary considerably. replication and segregation can occur from almost any position in the cell. Moreover, similar numbers of focus splitting and pairing events are observed in the population. This suggests a model for partition in which plasmids that are paired either as a result of replication or by fairly frequent chance encounters. They then move rapidly away from each other along the cell axis, maintaining a dynamic distribution of copies approximating an even distribution along the long cell axis. These observations have important implications for the mechanism for physical separation of the plasmids. This year, we have developed a mathematical simulation of plasmid segregation that resembles the type of model known as a Monte Carlo simulation. The simulation is derived from the rules for segregation that we have p[roposed to account for our experimental observations. The simulation produces plasmid segregation fidelity and spacial distributions of plasmid copies within the cell that closely resemble the experimental data in all details. The simulation greatly increases our confidence that the new picture of plasmid segregation that is emerging from our studies is correct.